A Smart Tea Garden Resource Dynamic Scheduling System and Method Integrating Environmental Monitoring

Abstract

This invention discloses a smart tea garden resource dynamic scheduling system and method integrating environmental monitoring, comprising the following steps: S1, collecting environmental and operational data, processing and generating a standardized dataset; S2, discrete modeling and generating an environmental field, topologically analyzing humidity and pest and disease intensity, identifying anomalies and triggering local scheduling; S3, generating a set of tasks, resources, and environment, constructing a task-resource-environment hypergraph, establishing a time-extended graph and forming constraints; S4, introducing dynamic game theory to generate scheduling schemes and control commands; S5, issuing commands to the execution end, collecting execution status and environmental data, and forming a feedback dataset; S6, constructing a two-layer memory scheduling unit to generate the parameter set for the next cycle. This invention, by integrating environmental monitoring and dynamic scheduling, achieves precise allocation and efficient management of tea garden resources, thereby improving production efficiency and ensuring ecological sustainability.

Description

Technical Field

[0001] This invention relates to the field of smart agriculture technology, and in particular to a smart tea garden resource dynamic scheduling system and method that integrates environmental monitoring. Background Technology

[0002] With the development of modern agriculture, especially the promotion of smart agriculture, farmland resource allocation based on information technology and automation has become a key means to improve agricultural production efficiency and sustainability. In tea garden management, the application of environmental monitoring technology has gradually become a core component, enabling real-time collection of multi-dimensional environmental data such as meteorology, soil, and pests and diseases, providing data support for scientific decision-making. Traditional tea garden resource allocation methods mainly rely on manual experience and static planning based on historical data, making it difficult to achieve real-time response and dynamic optimization. When faced with sudden environmental changes, traditional methods often cannot adjust the allocation plan in a timely manner, leading to resource waste or low production efficiency.

[0003] Currently, some existing tea garden resource scheduling systems combine environmental and operational data, but most are merely based on preset rules for fixed scheduling, lacking sufficient flexibility and adaptability, and unable to effectively cope with dynamically changing complex environments and task requirements. Furthermore, many existing systems suffer from information lag and slow response in the use of environmental monitoring data, resulting in the scheduling system failing to react promptly to emergencies such as pest and disease outbreaks and extreme weather, thus affecting the stability and yield of tea garden production.

[0004] The existing technology also has the following shortcomings: on the one hand, it lacks a comprehensive dynamic adjustment mechanism for changes in the tea garden environment and resource allocation, and cannot be optimized in real time based on real-time feedback; on the other hand, the existing scheduling system usually fails to effectively combine long-term ecological carrying capacity and short-term production goals, ignores the sustainability of ecological factors, resulting in inefficient use of resources and even a certain burden on the ecological environment.

[0005] Therefore, how to provide a smart tea garden resource dynamic scheduling system and method that integrates environmental monitoring is an urgent problem to be solved by those skilled in the art. Summary of the Invention

[0006] One objective of this invention is to propose a smart tea garden resource dynamic scheduling system and method that integrates environmental monitoring. This invention fully utilizes environmental monitoring, dynamic game optimization algorithms, and a two-layer memory scheduling unit technology. It details how to combine real-time environmental data and operational data to dynamically optimize task allocation and resource scheduling in tea garden resource scheduling. This method enhances the real-time response capability of tea garden resource scheduling by introducing a topological anomaly triggering mechanism, task-resource-environment hypergraph construction, and dynamic game solving process. The two-layer memory scheduling unit can continuously update and adjust based on execution feedback data, ensuring long-term sustainable resource use and ecological balance. This invention has the advantages of precise scheduling, efficient response, and ecological optimization, effectively addressing unexpected events in tea garden production while improving resource allocation efficiency and reducing energy consumption and environmental burden.

[0007] A method for dynamic scheduling of smart tea garden resources integrating environmental monitoring, according to an embodiment of the present invention, includes the following steps:

[0008] S1. Collect tea garden environment and operation data, process the tea garden environment and operation data, and construct standardized environmental dataset and operation dataset;

[0009] S2. Based on the standardized environmental dataset, the tea garden space is discretely modeled to generate an environmental field. Topological data analysis is performed on the humidity field and the pest and disease intensity field to extract topological features. Based on the preset hysteresis threshold and minimum duration, topological abnormal events are determined and local scheduling trigger orders are output.

[0010] S3. Generate a task set based on the job task information and local scheduling trigger order of the running dataset, generate a resource set based on the resource configuration information of the running dataset, generate an environment node set based on the environment field, construct a task-resource-environment hypergraph and establish a time-extended hypergraph, assign weights to the hyperedges, and form a constraint set.

[0011] S4. Introduce a dynamic game-solving process on the time-extended hypergraph to determine the resource-task-time allocation and obtain the scheduling scheme and control instruction set;

[0012] S5. Distribute the scheduling plan and control instruction set to pumps, valves, drones, transportation equipment and work teams for execution, and collect valve position, flow rate, energy consumption, task progress and environmental field updates in real time to form an execution feedback dataset;

[0013] S6. Based on the execution feedback dataset and the normalized environment dataset, establish a two-layer memory scheduling unit for resource-environment mutual feedback to generate a parameter set for the next scheduling cycle.

[0014] Optionally, the tea garden environment and operation data specifically include meteorological data, soil data, pest and disease image data, topographic information, irrigation network operation data, and operation task and resource allocation information.

[0015] Optionally, the processing of tea garden environment and operation data specifically includes performing noise reduction, missing value imputation, format standardization, and temporal and spatial alignment on the tea garden environment and operation data.

[0016] Optionally, S2 specifically includes:

[0017] S21. Based on the standardized environmental dataset, the tea garden space is gridded or plotted according to the rules along the tea row direction to generate a humidity field and a pest and disease intensity field synchronized with time, and spatial registration and temporal alignment are completed with the terrain information and irrigation network operation data.

[0018] S22. Construct multi-scale threshold sequences and sliding time windows in the humidity field and the pest and disease intensity field respectively, perform topological scanning on the field data under different spatial resolutions, extract the change trajectory of the number of connected branches and the number of loops, and extract the topological skeleton under the row direction constraint to form a candidate topological event set.

[0019] S23. Perform spatiotemporal persistence analysis on the candidate set of topological events, calculate the duration and spatial expansion of each candidate topological event, use the dual threshold hysteresis criterion and the minimum duration criterion for screening, and conduct consensus voting at the grid-plot-zone level to obtain a stable set of topological events.

[0020] S24. Perform cross-modal consistency check and hydraulic propagation calibration on the stable topological event set. The cross-modal consistency check requires that the humidity field and the pest and disease intensity field be consistent in location and evolution direction or meet the substitution consistency condition. The hydraulic propagation calibration determines the influence radius and effective time window of each stable topological event based on topographic information and irrigation network operation data.

[0021] S25. Generate local scheduling trigger orders according to the spatial location, influence radius, effective time window and evolution intensity of stable topological events.

[0022] Optionally, S3 specifically includes:

[0023] S31. Extract job task information from the running dataset to generate a task set, extract resource configuration information from the running dataset to generate a resource set, and generate an environmental node set from the environmental field. The environmental node includes spatial location, terrain information and hydraulic parameters.

[0024] S32. Construct a task-resource-environment hypergraph based on the task set, resource set, and environment node set. Establish a hyperedge for each task, connect the task with the required resource combination and related environment nodes, and map the local scheduling trigger order to the priority or hard time mark of the corresponding task or environment node.

[0025] S33. Establish a time-extended hypergraph, copy task, resource, and environment nodes at a preset scheduling interval, and allocate corresponding available time periods or effective time windows to task nodes, resource nodes, and environment nodes.

[0026] S34. Assign weights to the superedges according to the scheduling interval. The weighting comprehensively considers topological risk, terrain difficulty, water supply accessibility and feasible flow, resource arrival and movement time, equipment start-up and switching costs, energy consumption and operational safety constraints, and generates weights to characterize the comprehensive cost and priority of the scheduling period.

[0027] S35. Based on the interaction relationship between the task set, resource set, and environment node set in the time-extended hypergraph, extract the task time window constraint, resource capacity constraint, environment accessibility constraint, and hydraulic condition constraint, and revise the constraint conditions in combination with the hyperedge weights, and finally summarize to generate a constraint set.

[0028] Optionally, S4 specifically includes:

[0029] S41. Initialize the game-solving environment on the time-extended hypergraph and constraint set, set the task allocation indicator, task activation indicator, valve position control sequence and pump speed control sequence, load the hyperedge weight and local scheduling trigger command, and generate the infeasible region shielding graph.

[0030] S42. Execute a dual-channel weight update mechanism. The emergency channel sets hard time windows and increases priority for relevant task nodes and environment nodes based on local scheduling trigger orders. The regular channel updates the super-edge weights on a rolling basis based on resource arrival time, equipment switching cost, energy consumption and operation safety constraints. It outputs time period priority sequence, priority mapping table and hard time window list, and merges them with the infeasible region shielding graph to form a constraint enhancement set.

[0031] S43. Initiate the optimal response game process driven by shadow prices, introduce shadow price parameters corresponding to ecological budget and water supply budget, execute optimal response updates in the order of resource clusters, and use the previous round of decisions for resource clusters that have not been updated. Simultaneously generate valve group linkage control sequence, set minimum holding time and silent interval, and output initial scheduling scheme, valve group linkage control sequence and stable interval list.

[0032] S44. When a new topology anomaly or external disturbance is received, a local re-optimization is triggered within the constraint enhancement set. The issued instructions within the stable interval remain unchanged. Only the task allocation, equipment path, valve position sequence and pump speed sequence of the time period that has not been executed are incrementally corrected. An upper limit is set on the change range. The corrected scheduling scheme and control instruction candidate set are output.

[0033] S45. Generate the final scheduling scheme and control instruction set based on the revised scheduling scheme and control instruction candidate set, and generate the accompanying annotation set at the same time.

[0034] Optionally, S5 specifically includes:

[0035] S51. Distribute the scheduling plan and control instruction set to the terminals of pumps, valves, drones, transportation equipment and work teams, along with time windows, priorities, allowable deviations and rollback plans;

[0036] S52. Perform pre-execution verification on control commands. The verification includes hydraulic safety, path accessibility, energy consumption budget and operational safety. Control commands that do not meet the conditions are marked as delayed and returned to the solver. Control commands that meet the conditions are written into the execution queue.

[0037] S53. Execute control commands according to the time window and sequence specified in the scheduling plan, drive valve opening and closing, pump speed adjustment, UAV flight path, transportation equipment path and work order of the work team, and generate execution status record;

[0038] S54. During execution, real-time data collection is performed on valve position, flow rate, pressure, energy consumption, equipment operating status, position and speed, task completion, abnormal events, and environmental field updates. An execution log is generated based on a unified time reference, and safety measures are triggered according to the rollback plan when a sudden abnormal event occurs.

[0039] S55. Process the execution logs, including noise reduction, completion, standardization, and time alignment, and summarize them to form an execution feedback dataset.

[0040] Optionally, S6 specifically includes:

[0041] S61. Construct a two-layer memory scheduling unit, which includes a short-term memory module, a long-term memory module, a rule engine and a parameter management module. The short-term memory module establishes a time-series update mechanism based on sliding window, exponentially weighted moving average and online change point detection. The long-term memory module establishes a periodic learning mechanism based on seasonal-trend decomposition, clustering and exponential smoothing.

[0042] S62. Extract valve position execution delay, pump speed deviation, flow deviation, energy consumption deviation, task delay, path offset, environmental field fitting residual and trigger response lag from the execution feedback dataset, perform time series merging, denoising and time alignment, generate short-term deviation vector, and submit it to the rule engine.

[0043] S63. Based on the short-term deviation vector, the rule engine completes online revision. The revision content includes the topology anomaly judgment hysteresis threshold and minimum duration, the topology risk component, terrain difficulty component and hydraulic feasibility component in the superedge weight, the minimum holding time for valve position switching, the resource movement cost coefficient and the pre-execution verification threshold, forming a short-term parameter set.

[0044] S64. Aggregate the execution feedback dataset and the normalized environment dataset according to the preset cycle, extract the rainfall-evapotranspiration phase, pest and disease diffusion phase and operation efficiency curve, complete the seasonal-trend decomposition and clustering archive, revise the ecological and resource scheduling parameters based on exponential smoothing, generate the shadow price parameter range and scheduling steady state threshold, and form a long-term parameter set.

[0045] S65. The parameter management module merges the short-term parameter set and the long-term parameter set, performs consistency verification and security verification, and generates the parameter set for the next scheduling cycle.

[0046] A smart tea garden resource dynamic scheduling system integrating environmental monitoring according to an embodiment of the present invention includes the following modules:

[0047] The data acquisition and processing module is used to collect and preprocess environmental and operational data to generate standardized environmental datasets and operational datasets.

[0048] The topology anomaly analysis module is used to analyze humidity and pest and disease intensity, extract topological features and identify anomalies, and output local scheduling trigger orders.

[0049] The hypergraph construction module is used to construct the task-resource-environment hypergraph and the time-extended hypergraph, and to assign weights to hyperedges and form constraint sets.

[0050] The dynamic game scheduling module is used to perform game solving on the time-extended hypergraph and constraint set, and output scheduling scheme and control instruction set;

[0051] The execution control and feedback module is used to issue scheduling scheme combination control instruction sets, collect execution status and environment updates, and generate execution feedback datasets.

[0052] The two-layer memory scheduling unit is used to update short-term thresholds and weights, revise long-term ecological budgets and resource priors, and output the parameter set for the next scheduling cycle.

[0053] The beneficial effects of this invention are:

[0054] This invention achieves a more precise and flexible scheduling strategy in the dynamic allocation of tea garden resources by integrating environmental monitoring and dynamic game optimization algorithms. Compared with traditional tea garden management methods, this invention can collect and analyze multi-source environmental data in real time and dynamically adjust the scheduling plan based on task assignments and resource status. By introducing a topology anomaly triggering mechanism, the system can respond promptly to sudden environmental changes, avoiding the slow response of traditional systems to sudden events such as extreme weather or pests and diseases, and significantly improving the emergency response capability of tea garden resource management.

[0055] This invention utilizes a dual-layer memory scheduling unit, combining short-term and long-term memory update mechanisms, enabling the system to not only rapidly respond to immediate changes but also perform long-term ecological optimization and resource prediction. The short-term memory module adjusts the topology anomaly detection threshold and hyperedge weights in real time, ensuring the accuracy and timeliness of scheduling; the long-term memory module updates resource capabilities and ecological budgets based on historical data and ecological patterns, ensuring the sustainability of tea garden resources in long-term operation and enhancing the adaptive capability and optimization effect of the tea garden scheduling system.

[0056] By optimizing resource allocation and task scheduling, this invention can reduce resource waste and energy consumption. In the long run, this invention helps maintain ecological balance, reduce problems such as excessive water pumping and excessive fertilization, protect the ecological environment of tea gardens, and enhance the overall sustainable development capacity of tea gardens. Attached Figure Description

[0057] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0058] Figure 1 The flowchart shows a smart tea garden resource dynamic scheduling method that integrates environmental monitoring, as proposed in this invention.

[0059] Figure 2 This is a schematic diagram of the structure of a smart tea garden resource dynamic scheduling system that integrates environmental monitoring, as proposed in this invention. Detailed Implementation

[0060] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, illustrating only the basic structure of the invention, and therefore only show the components relevant to the invention.

[0061] refer to Figure 1 A method for dynamic scheduling of smart tea garden resources that integrates environmental monitoring includes the following steps:

[0062] S1. Collect tea garden environment and operation data, process the tea garden environment and operation data, and construct standardized environmental dataset and operation dataset;

[0063] S2. Based on the standardized environmental dataset, the tea garden space is discretely modeled to generate an environmental field. Topological data analysis is performed on the humidity field and the pest and disease intensity field to extract topological features. Based on the preset hysteresis threshold and minimum duration, topological abnormal events are determined and local scheduling trigger orders are output.

[0064] S3. Generate a task set based on the job task information and local scheduling trigger order of the running dataset, generate a resource set based on the resource configuration information of the running dataset, generate an environment node set based on the environment field, construct a task-resource-environment hypergraph and establish a time-extended hypergraph, assign weights to the hyperedges, and form a constraint set.

[0065] S4. Introduce a dynamic game-solving process on the time-extended hypergraph to determine the resource-task-time allocation and obtain the scheduling scheme and control instruction set;

[0066] S5. Distribute the scheduling plan and control instruction set to pumps, valves, drones, transportation equipment and work teams for execution, and collect valve position, flow rate, energy consumption, task progress and environmental field updates in real time to form an execution feedback dataset;

[0067] S6. Based on the execution feedback dataset and the normalized environment dataset, establish a two-layer memory scheduling unit for resource-environment mutual feedback to generate a parameter set for the next scheduling cycle.

[0068] In this embodiment, the tea garden environment and operation data specifically include meteorological data, soil data, pest and disease image data, topographic information, irrigation network operation data, and operation task and resource allocation information.

[0069] In this embodiment, the processing of tea garden environment and operation data specifically includes performing noise reduction, missing value imputation, format standardization, and temporal and spatial alignment on the tea garden environment and operation data.

[0070] In this embodiment, S2 specifically includes:

[0071] S21. Based on the standardized environmental dataset, the tea garden space is gridded or plotted according to the rules along the tea row direction to generate a humidity field and a pest and disease intensity field synchronized with time, and spatial registration and temporal alignment are completed with the terrain information and irrigation network operation data.

[0072] S22. Construct multi-scale threshold sequences and sliding time windows in the humidity field and the pest and disease intensity field respectively, perform topological scanning on the field data under different spatial resolutions, extract the change trajectory of the number of connected branches and the number of loops, and extract the topological skeleton under the row direction constraint to form a candidate topological event set.

[0073] S23. Perform spatiotemporal persistence analysis on the candidate set of topological events, calculate the duration and spatial expansion of each candidate topological event, and use a dual-threshold hysteresis criterion and a minimum duration criterion for screening. Consistency voting is then performed at the grid-plot-zone three-level to obtain a stable set of topological events. Specifically, the calculation of the duration and spatial expansion of each candidate topological event is as follows:

[0074] The start and end times of each candidate topological event in continuous time series data are statistically analyzed, and the total time interval in which the candidate topological event exists is calculated as the duration.

[0075] In the spatial distribution of the same candidate topological event, the coverage or area change of the candidate topological event on the spatial grid point or plot is statistically analyzed to obtain the maximum spatial expansion.

[0076] By combining time series and spatial distribution, a "time-space" trajectory is formed for each candidate event, the dynamic changes in spatial range during the duration are analyzed, and the maximum spatiotemporal impact range is extracted as an evaluation index.

[0077] Among them, the dual-threshold hysteresis criterion refers to using two thresholds, high and low, to determine the generation and decay of candidate topological events. An event is considered to have occurred only when the monitoring index exceeds the upper threshold, and the event is considered to have ended only when it falls below the lower threshold. The minimum duration criterion refers to the fact that a candidate topological event is only considered a valid event when its duration reaches the preset minimum duration requirement, thus filtering out short-term anomalies and noise.

[0078] S24. Perform cross-modal consistency check and hydraulic propagation calibration on the stable topological event set. The cross-modal consistency check requires that the humidity field and the pest and disease intensity field be consistent in location and evolution direction or meet the substitution consistency condition. The hydraulic propagation calibration determines the influence radius and effective time window of each stable topological event based on topographic information and irrigation network operation data.

[0079] S25. Generate a local scheduling trigger order according to the spatial location, influence radius, effective time window and evolution intensity of the stable topology event. The local scheduling trigger order includes location identifier, influence radius, priority, limited operation type, start and end time window and reliability level.

[0080] This invention introduces topology analysis, multi-scale thresholds, and spatiotemporal persistence criteria into the tea garden environmental monitoring and scheduling process, achieving high-precision identification and dynamic response to humidity and pest / disease anomalies. It innovatively employs multi-level screening of "candidate topology events," dual-threshold hysteresis, and a minimum duration mechanism to enhance the system's robustness to short-term noise and sporadic anomalies, avoiding false alarms and missed alarms. Cross-modal consistency verification and hydraulic propagation calibration further ensure the scientific nature of anomaly judgment and the accuracy of scheduling response. Compared to traditional methods relying on manual experience or single threshold judgment, this invention can identify sudden disease and moisture risks in real time, intelligently, and accurately, and automatically generate scheduling strategies based on zones, times, and intensities. This improves the emergency response capability and resource allocation efficiency of tea gardens, effectively ensuring the healthy growth of tea trees and ecological sustainability, demonstrating extremely high innovation and practical value.

[0081] In this embodiment, S3 specifically includes:

[0082] S31. Extract job task information from the running dataset to generate a task set, extract resource configuration information from the running dataset to generate a resource set, and generate an environmental node set from the environmental field. The environmental node includes spatial location, terrain information and hydraulic parameters.

[0083] S32. Construct a task-resource-environment hypergraph based on the task set, resource set, and environment node set. Establish a hyperedge for each task, connect the task with the required resource combination and related environment nodes, and map the local scheduling trigger order to the priority or hard time mark of the corresponding task or environment node.

[0084] S33. Establish a time-extended hypergraph, copy task, resource, and environment nodes at a preset scheduling interval, and allocate corresponding available time periods or effective time windows to task nodes, resource nodes, and environment nodes.

[0085] S34. Assign weights to the superedges according to the scheduling interval. The weighting comprehensively considers topological risk, terrain difficulty, water supply accessibility and feasible flow, resource arrival and movement time, equipment start-up and switching costs, energy consumption and operational safety constraints, and generates weights to characterize the comprehensive cost and priority of the scheduling period.

[0086] S35. Based on the interaction relationships between the task set, resource set, and environment node set in the time-extended hypergraph, extract task time window constraints, resource capacity constraints, environmental accessibility constraints, and hydraulic condition constraints. Combine these constraints with hyperedge weights to revise the restrictions and finally generate a constraint set. The constraint set includes resource capacity and concurrency limit, unique activation and full allocation of tasks within the time window, consistency between resource allocation and task activation, accessibility based on slope and slipperiness level, hydraulic feasibility based on pressure, flow rate, and branch status, safety distance between personnel and equipment and operational interference restrictions, valve position switching frequency and minimum holding time, and path avoidance and restricted area restrictions.

[0087] In this embodiment, S4 specifically includes:

[0088] S41. Initialize the game-solving environment on the time-extended hypergraph and constraint set, set the task allocation indicator, task activation indicator, valve position control sequence and pump speed control sequence, load the hyperedge weight and local scheduling trigger command, and generate the infeasible region shielding graph.

[0089] S42. Execute a dual-channel weight update mechanism. The emergency channel sets hard time windows and increases priority for relevant task nodes and environment nodes based on local scheduling trigger orders. The regular channel updates the super-edge weights on a rolling basis based on resource arrival time, equipment switching cost, energy consumption, and operational safety constraints. Output a time period priority sequence, priority mapping table, and hard time window list, and merge them with the infeasible region masking graph to form a constraint enhancement set, where:

[0090] The emergency channel is designed to quickly prioritize relevant tasks and environmental nodes and set hard time windows for sudden and abnormal events.

[0091] The regular channel is used for daily scheduling tasks, and the weights are dynamically optimized based on resources, energy consumption, and security to achieve hierarchical management of scheduling responses;

[0092] S43. Initiate the optimal response game process driven by shadow prices, introduce shadow price parameters corresponding to ecological budget and water supply budget, execute optimal response updates in the order of resource clusters, and use the previous round of decisions for resource clusters that have not been updated. Simultaneously generate valve group linkage control sequence, set minimum hold time and silent interval, and output initial scheduling scheme, valve group linkage control sequence and stable interval list. Among them, shadow price refers to the virtual price used to measure the impact of each unit increase or decrease of resources on the overall scheduling target in resource scheduling optimization, ecological budget refers to the upper limit of resource consumption set to ensure ecological balance, and water supply budget refers to the total amount of irrigation water set according to the availability of water resources.

[0093] S44. When a new topology anomaly or external disturbance is received, a local re-optimization is triggered within the constraint enhancement set. The issued instructions within the stable interval remain unchanged. Only the task allocation, equipment path, valve position sequence and pump speed sequence of the time period that has not been executed are incrementally corrected. An upper limit is set on the change range. The corrected scheduling scheme and control instruction candidate set are output.

[0094] S45. Generate the final scheduling scheme and control instruction set based on the revised scheduling scheme and control instruction candidate set, and generate an accompanying annotation set. The accompanying annotation set records the trigger source, applicable time window, priority, allowable deviation, verification conditions, rollback scheme and safety verification results.

[0095] This invention achieves efficient hierarchical management and flexible response in the scheduling process by introducing a time-extended hypergraph and a hierarchical dynamic weight update mechanism, enabling separate optimization for sudden anomalies and routine tasks. It innovatively integrates parameters such as shadow prices, ecological budgets, and water supply budgets into the game-theoretic solution process, making resource allocation more flexible and forward-looking, effectively balancing ecological protection and production efficiency. A local re-optimization mechanism ensures the system's sustainable and stable operation under emergencies, avoiding frequent changes to issued instructions and enabling dynamic correction and optimization of the scheduling scheme. Compared with traditional static scheduling or single-priority strategies, this scheme significantly improves the emergency response capability, resource allocation efficiency, and ecological sustainability of tea garden management, demonstrating excellent intelligent decision-making capabilities and practical application value in complex scenarios with multiple objectives and constraints.

[0096] In this embodiment, S5 specifically includes:

[0097] S51. Distribute the scheduling plan and control instruction set to the terminals of pumps, valves, drones, transportation equipment and work teams, along with time windows, priorities, allowable deviations and rollback plans;

[0098] S52. Perform pre-execution verification on control commands. The verification includes hydraulic safety, path accessibility, energy consumption budget and operational safety. Control commands that do not meet the conditions are marked as delayed and returned to the solver. Control commands that meet the conditions are written into the execution queue.

[0099] S53. Execute control commands according to the time window and sequence specified in the scheduling plan, drive valve opening and closing, pump speed adjustment, UAV flight path, transportation equipment path and work order of the work team, and generate execution status record;

[0100] S54. During execution, real-time data collection is performed on valve position, flow rate, pressure, energy consumption, equipment operating status, position and speed, task completion, abnormal events, and environmental field updates. An execution log is generated based on a unified time reference, and safety measures are triggered according to the rollback plan when a sudden abnormal event occurs.

[0101] S55. Process the execution logs, including noise reduction, completion, standardization, and time alignment, and summarize them to form an execution feedback dataset. The execution feedback dataset contains device-level, task-level, and environment-level records, and establishes traceability associations with scheduling schemes and control instruction sets.

[0102] In this embodiment, S6 specifically includes:

[0103] S61. Construct a two-layer memory scheduling unit, which includes a short-term memory module, a long-term memory module, a rule engine and a parameter management module. The short-term memory module establishes a time-series update mechanism based on sliding window, exponentially weighted moving average and online change point detection. The long-term memory module establishes a periodic learning mechanism based on seasonal-trend decomposition, clustering and exponential smoothing.

[0104] S62. Extract valve position execution delay, pump speed deviation, flow deviation, energy consumption deviation, task delay, path offset, environmental field fitting residual and trigger response lag from the execution feedback dataset, perform time series merging, denoising and time alignment, generate short-term deviation vector, and submit it to the rule engine.

[0105] S63. Based on the short-term deviation vector, the rule engine completes online revision. The revision content includes the topology anomaly judgment hysteresis threshold and minimum duration, the topology risk component, terrain difficulty component and hydraulic feasibility component in the superedge weight, the minimum holding time for valve position switching, the resource movement cost coefficient and the pre-execution verification threshold, forming a short-term parameter set.

[0106] S64. Aggregate the execution feedback dataset and the normalized environment dataset according to a preset cycle, extract the rainfall-evapotranspiration phase, pest and disease diffusion phase, and operation efficiency curve, complete seasonal-trend decomposition and clustering archiving, revise the ecological and resource scheduling parameters based on exponential smoothing, generate shadow price parameter ranges and scheduling steady-state thresholds, and form a long-term parameter set. The specific steps of revising the ecological and resource scheduling parameters based on exponential smoothing are as follows:

[0107] By using exponential smoothing to weight the historical series of rainfall and evapotranspiration, the long-term equilibrium level of water supply and ecological consumption is updated, avoiding interference from short-term extreme fluctuations on parameter settings.

[0108] The feedback data, including equipment energy consumption and operating efficiency, are subjected to exponential smoothing, and the prior parameters of resource capacity are revised accordingly.

[0109] The operation and maintenance parameters, including restricted access rules, maintenance windows and inspection frequencies, are smoothed by applying exponential smoothing and combined with seasonal fluctuations to form a smoothed revision result, generating a stable shadow price parameter range and a scheduling steady-state threshold.

[0110] S65. The parameter management module merges the short-term parameter set and the long-term parameter set, performs consistency verification and security verification, generates the parameter set for the next scheduling cycle, and marks it with version number, effective time, parameter scope, and rollback conditions. The specific steps of performing consistency verification and security verification are as follows:

[0111] The short-term parameter set is compared with the long-term parameter set to check for conflicts or mismatches in threshold setting, weight update and resource capability priors; if differences are found, revisions are made by weighted averaging.

[0112] Based on the parameter set after consistency verification, each parameter is verified to ensure that it meets the requirements of hydraulic safety, equipment operation restrictions, safe working distance, and energy consumption limit. Parameters that do not meet the safety requirements are adjusted or rolled back to ensure that the generated scheduling parameters are safe and reliable in implementation.

[0113] This invention constructs a dual-layer memory scheduling unit, organically combining short-term deviation correction with long-term trend learning to form a dynamic and periodic parallel parameter update mechanism, thereby improving the system's adaptability and stability. The short-term memory module, relying on sliding windows, weighted averaging, and change point detection, can quickly capture immediate anomalies such as valve position delays, energy consumption deviations, and task delays, enabling online revision of scheduling thresholds and weights. The long-term memory module, based on seasonal-trend decomposition and exponential smoothing, learns from the periodic patterns of rainfall, pest and disease spread, and operational efficiency, generating robust shadow price ranges and scheduling thresholds. The innovation lies in simultaneously considering immediate response and long-term evolution, enabling the parameter set to quickly adapt to environmental disturbances while reflecting ecological balance and resource carrying capacity, thus avoiding the problems of parameter solidification, lag, or imbalance in traditional scheduling. Finally, the system performs consistency and security checks in each cycle to ensure that the generated parameters are both reasonable and safe, achieving a comprehensive effect of improving scheduling accuracy, reducing energy and water consumption, and enhancing ecological sustainability.

[0114] refer to Figure 2 A smart tea garden resource dynamic scheduling system integrating environmental monitoring includes the following modules:

[0115] The data acquisition and processing module is used to collect and preprocess environmental and operational data to generate standardized environmental datasets and operational datasets.

[0116] The topology anomaly analysis module is used to analyze humidity and pest and disease intensity, extract topological features and identify anomalies, and output local scheduling trigger orders.

[0117] The hypergraph construction module is used to construct the task-resource-environment hypergraph and the time-extended hypergraph, and to assign weights to hyperedges and form constraint sets.

[0118] The dynamic game scheduling module is used to perform game solving on the time-extended hypergraph and constraint set, and output scheduling scheme and control instruction set;

[0119] The execution control and feedback module is used to issue scheduling scheme combination control instruction sets, collect execution status and environment updates, and generate execution feedback datasets.

[0120] The two-layer memory scheduling unit is used to update short-term thresholds and weights, revise long-term ecological budgets and resource priors, and output the parameter set for the next scheduling cycle.

[0121] Example 1:

[0122] To verify the feasibility of this invention in practice, it was applied to a high-altitude tea garden (approximately 120 acres). This tea garden has complex terrain with significant differences in slope between the north and south sides, leading to common problems such as uneven water and fertilizer distribution and localized outbreaks of pests and diseases. Traditional irrigation methods rely on manual experience, often resulting in the following issues: severe waterlogging in low-lying areas causing root hypoxia in the tea trees; insufficient irrigation in high-slope areas leading to slow shoot development; and delayed manual intervention when leafhoppers or anthracnose occur, resulting in reduced yields.

[0123] In practical applications, the system first collects environmental data through deployed weather stations, soil moisture sensors, insect monitoring lamps, and cameras, and then performs noise reduction, missing value imputation, and spatiotemporal alignment processing on the data. From May 8th to May 14th, 2024, the tea garden experienced continuous rainy weather, with an average humidity of around 87%. The soil moisture content in the 0–20 cm depth of the low-lying area on the south slope reached 36%, far exceeding the suitable threshold (around 25%). At the same time, drone image monitoring showed that the proportion of leaf spots on the tea leaves on the south slope increased from 3% to 9%.

[0124] The system detected persistent anomalies in the humidity and disease intensity fields in the topology anomaly analysis module. Topological skeleton analysis identified the low-lying area on the south slope as a high-risk point, triggering a local scheduling command. The scheduling module immediately generated a task-resource-environment hypergraph and used a game theory approach on the time-extended hypergraph, incorporating constraints: irrigation was stopped on the south slope, while a low-flow water supply was maintained on the higher slopes; a drone was scheduled to conduct localized spraying on the morning of May 10th, using pyraclostrobin as the pesticide, covering an area of ​​20 mu (approximately 1.3 hectares); the work team was reassigned to the south slope for manual inspection and pruning of diseased leaves.

[0125] During the execution phase, the system issued control commands: close the valves of the three irrigation branch lines on the south slope, adjust the water supply flow on the north slope to 12 m³ per hour, and have drones complete the spraying operation from 7:30 to 9:00 on May 10th. The system simultaneously collected real-time data on valve positions, flow rates, energy consumption, and disease expansion, and revised the system according to the rollback plan in the event of abnormal events.

[0126] Feedback data showed that on May 12, the soil moisture content on the south slope recovered to 27%, and the proportion of lesions remained stable at around 7%, without further expansion; the number of tea green leafhoppers captured decreased by 28% compared to May 9. Through updates from the dual-layer memory scheduling unit, the valve switching holding time, disease threshold, and energy consumption deviation data generated during this scheduling were recorded and used to guide subsequent cycles.

[0127] Table 1. Environmental Monitoring and Scheduling Execution Data of Tea Garden

[0128]

[0129] As shown in Table 1, the average air humidity in the tea garden remained between 72% and 89% during the continuous rainy weather, significantly higher than the suitable standard of around 70%. The soil moisture content in the low-lying areas on the south slope reached 34% on May 9th and rose to 36% on May 10th, exceeding the reasonable threshold of approximately 25%, directly triggering the system's topological anomaly warning. With the increase in humidity, the proportion of lesion area gradually expanded from 3% on May 8th to 9% on May 11th, and the number of tea green leafhoppers captured also peaked at 20 per day on May 10th. These indicators all reflect a rapid increase in the risk of pest and disease outbreaks.

[0130] After system intervention, a differentiated irrigation strategy of "closing the branch road on the south slope and supplying water at low flow rates on the high slope" was adopted. On May 10, a drone was used to spray pyraclostrobin on a 20-mu plot on the south slope, and the number of manual inspections was increased to 3 times. The results showed that the soil moisture content had recovered to 27% by May 12, the proportion of lesions had decreased to 7%, and the number of tea green leafhoppers had decreased to 14 per day, indicating that the spread of the disease had been effectively controlled.

[0131] From a resource consumption perspective, daily water consumption decreased from 180–190 tons before the scheduling to a minimum of 110 tons on May 11th, demonstrating significant water conservation. Energy consumption also decreased accordingly with the control of valves and reduction of pump operation, from 95–98 kWh to 70–78 kWh. The tea yield reduction rate remained within 0%–2% throughout the process, compared to the potential yield reduction of over 8% that could occur with traditional methods, demonstrating clear economic and ecological benefits.

[0132] Comprehensive analysis shows that the system of the present invention achieves the goals of "timely loss mitigation, water and energy conservation, and guaranteed yield" by accurately identifying topological anomalies and dynamically adjusting irrigation and disease control tasks in high humidity environments, fully demonstrating the feasibility and effectiveness of the present invention in real tea garden management.

[0133] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for dynamic scheduling of smart tea garden resources integrating environmental monitoring, characterized in that, Includes the following steps: S1. Collect tea garden environment and operation data, process the tea garden environment and operation data, and construct standardized environmental dataset and operation dataset; S2. Based on the standardized environmental dataset, the tea garden space is discretely modeled to generate an environmental field. Topological data analysis is performed on the humidity field and the pest and disease intensity field to extract topological features. Based on the preset hysteresis threshold and minimum duration, topological abnormal events are determined and local scheduling trigger orders are output. S3. Generate a task set based on the job task information and local scheduling trigger order of the running dataset, generate a resource set based on the resource configuration information of the running dataset, generate an environment node set based on the environment field, construct a task-resource-environment hypergraph and establish a time-extended hypergraph, assign weights to the hyperedges, and form a constraint set. S4. Introduce a dynamic game-solving process on the time-extended hypergraph to determine the resource-task-time allocation and obtain the scheduling scheme and control instruction set; S5. Distribute the scheduling plan and control instruction set to pumps, valves, drones, transportation equipment and work teams for execution, and collect valve position, flow rate, energy consumption, task progress and environmental field updates in real time to form an execution feedback dataset; S6. Based on the execution feedback dataset and the normalized environment dataset, establish a two-layer memory scheduling unit for resource-environment mutual feedback to generate a parameter set for the next scheduling cycle.

2. The method for dynamic scheduling of smart tea garden resources integrating environmental monitoring according to claim 1, characterized in that, The tea garden environment and operation data specifically include meteorological data, soil data, pest and disease image data, topographic information, irrigation network operation data, and operation tasks and resource allocation information.

3. The method for dynamic scheduling of smart tea garden resources integrating environmental monitoring as described in claim 1, characterized in that, The specific processing of tea garden environment and operation data includes denoising, missing value imputation, format standardization, and time-space alignment of the tea garden environment and operation data.

4. The method for dynamic scheduling of smart tea garden resources integrating environmental monitoring as described in claim 1, characterized in that, S2 specifically includes: S21. Based on the standardized environmental dataset, the tea garden space is gridded or plotted according to the rules along the tea row direction to generate a humidity field and a pest and disease intensity field synchronized with time, and spatial registration and temporal alignment are completed with the terrain information and irrigation network operation data. S22. Construct multi-scale threshold sequences and sliding time windows in the humidity field and the pest and disease intensity field respectively, perform topological scanning on the field data under different spatial resolutions, extract the change trajectory of the number of connected branches and the number of loops, and extract the topological skeleton under the row direction constraint to form a candidate topological event set. S23. Perform spatiotemporal persistence analysis on the candidate set of topological events, calculate the duration and spatial expansion of each candidate topological event, use the dual threshold hysteresis criterion and the minimum duration criterion for screening, and conduct consensus voting at the grid-plot-zone level to obtain a stable set of topological events. S24. Perform cross-modal consistency check and hydraulic propagation calibration on the stable topological event set. The cross-modal consistency check requires that the humidity field and the pest and disease intensity field be consistent in location and evolution direction or meet the substitution consistency condition. The hydraulic propagation calibration determines the influence radius and effective time window of each stable topological event based on topographic information and irrigation network operation data. S25. Generate local scheduling trigger orders according to the spatial location, influence radius, effective time window and evolution intensity of stable topological events.

5. The method for dynamic scheduling of smart tea garden resources integrating environmental monitoring as described in claim 1, characterized in that, S3 specifically includes: S31. Extract job task information from the running dataset to generate a task set, extract resource configuration information from the running dataset to generate a resource set, and generate an environmental node set from the environmental field. The environmental node includes spatial location, terrain information and hydraulic parameters. S32. Construct a task-resource-environment hypergraph based on the task set, resource set, and environment node set. Establish a hyperedge for each task, connect the task with the required resource combination and related environment nodes, and map the local scheduling trigger order to the priority or hard time mark of the corresponding task or environment node. S33. Establish a time-extended hypergraph, copy task, resource, and environment nodes at a preset scheduling interval, and allocate corresponding available time periods or effective time windows to task nodes, resource nodes, and environment nodes. S34. Assign weights to the superedges according to the scheduling interval. The weighting comprehensively considers topological risk, terrain difficulty, water supply accessibility and feasible flow, resource arrival and movement time, equipment start-up and switching costs, energy consumption and operational safety constraints, and generates weights to characterize the comprehensive cost and priority of the scheduling period. S35. Based on the interaction relationship between the task set, resource set, and environment node set in the time-extended hypergraph, extract the task time window constraint, resource capacity constraint, environment accessibility constraint, and hydraulic condition constraint, and revise the constraint conditions in combination with the hyperedge weights, and finally summarize to generate a constraint set.

6. The method for dynamic scheduling of smart tea garden resources integrating environmental monitoring according to claim 1, characterized in that, S4 specifically includes: S41. Initialize the game-solving environment on the time-extended hypergraph and constraint set, set the task allocation indicator, task activation indicator, valve position control sequence and pump speed control sequence, load the hyperedge weight and local scheduling trigger command, and generate the infeasible region shielding graph. S42. Execute a dual-channel weight update mechanism. The emergency channel sets hard time windows and increases priority for relevant task nodes and environment nodes based on local scheduling trigger orders. The regular channel updates the super-edge weights on a rolling basis based on resource arrival time, equipment switching cost, energy consumption and operation safety constraints. It outputs time period priority sequence, priority mapping table and hard time window list, and merges them with the infeasible region shielding graph to form a constraint enhancement set. S43. Initiate the optimal response game process driven by shadow prices, introduce shadow price parameters corresponding to ecological budget and water supply budget, execute optimal response updates in the order of resource clusters, and use the previous round of decisions for resource clusters that have not been updated. Simultaneously generate valve group linkage control sequence, set minimum holding time and silent interval, and output initial scheduling scheme, valve group linkage control sequence and stable interval list. S44. When a new topology anomaly or external disturbance is received, a local re-optimization is triggered within the constraint enhancement set. The issued instructions within the stable interval remain unchanged. Only the task allocation, equipment path, valve position sequence and pump speed sequence of the time period that has not been executed are incrementally corrected. An upper limit is set on the change range. The corrected scheduling scheme and control instruction candidate set are output. S45. Generate the final scheduling scheme and control instruction set based on the revised scheduling scheme and control instruction candidate set, and generate the accompanying annotation set at the same time.

7. The method for dynamic scheduling of smart tea garden resources integrating environmental monitoring according to claim 1, characterized in that, S5 specifically includes: S51. Distribute the scheduling plan and control instruction set to the terminals of pumps, valves, drones, transportation equipment and work teams, along with time windows, priorities, allowable deviations and rollback plans; S52. Perform pre-execution verification on control commands. The verification includes hydraulic safety, path accessibility, energy consumption budget and operational safety. Control commands that do not meet the conditions are marked as delayed and returned to the solver. Control commands that meet the conditions are written into the execution queue. S53. Execute control commands according to the time window and sequence specified in the scheduling plan, drive valve opening and closing, pump speed adjustment, UAV flight path, transportation equipment path and work order of the work team, and generate execution status record; S54. During execution, real-time data collection is performed on valve position, flow rate, pressure, energy consumption, equipment operating status, position and speed, task completion, abnormal events, and environmental field updates. An execution log is generated based on a unified time reference, and safety measures are triggered according to the rollback plan when a sudden abnormal event occurs. S55. Process the execution logs, including noise reduction, completion, standardization, and time alignment, and summarize them to form an execution feedback dataset.

8. The method for dynamic scheduling of smart tea garden resources integrating environmental monitoring according to claim 1, characterized in that, S6 specifically includes: S61. Construct a two-layer memory scheduling unit, which includes a short-term memory module, a long-term memory module, a rule engine and a parameter management module. The short-term memory module establishes a time-series update mechanism based on sliding window, exponentially weighted moving average and online change point detection. The long-term memory module establishes a periodic learning mechanism based on seasonal-trend decomposition, clustering and exponential smoothing. S62. Extract valve position execution delay, pump speed deviation, flow deviation, energy consumption deviation, task delay, path offset, environmental field fitting residual and trigger response lag from the execution feedback dataset, perform time series merging, denoising and time alignment, generate short-term deviation vector, and submit it to the rule engine. S63. Based on the short-term deviation vector, the rule engine completes online revision. The revision content includes the topology anomaly judgment hysteresis threshold and minimum duration, the topology risk component, terrain difficulty component and hydraulic feasibility component in the superedge weight, the minimum holding time for valve position switching, the resource movement cost coefficient and the pre-execution verification threshold, forming a short-term parameter set. S64. Aggregate the execution feedback dataset and the normalized environment dataset according to the preset cycle, extract the rainfall-evapotranspiration phase, pest and disease diffusion phase and operation efficiency curve, complete the seasonal-trend decomposition and clustering archive, revise the ecological and resource scheduling parameters based on exponential smoothing, generate the shadow price parameter range and scheduling steady state threshold, and form a long-term parameter set. S65. The parameter management module merges the short-term parameter set and the long-term parameter set, performs consistency verification and security verification, and generates the parameter set for the next scheduling cycle.

9. A smart tea garden resource dynamic scheduling system integrating environmental monitoring, comprising executing the smart tea garden resource dynamic scheduling method integrating environmental monitoring as described in any one of claims 1 to 8, characterized in that, Includes the following modules: The data acquisition and processing module is used to collect and preprocess environmental and operational data to generate standardized environmental datasets and operational datasets. The topology anomaly analysis module is used to analyze humidity and pest and disease intensity, extract topological features and identify anomalies, and output local scheduling trigger orders. The hypergraph construction module is used to construct the task-resource-environment hypergraph and the time-extended hypergraph, and to assign weights to hyperedges and form constraint sets. The dynamic game scheduling module is used to perform game solving on the time-extended hypergraph and constraint set, and output scheduling scheme and control instruction set; The execution control and feedback module is used to issue scheduling schemes and control instruction sets, collect execution status and environment updates, and generate execution feedback datasets. The two-layer memory scheduling unit is used to update short-term thresholds and weights, revise long-term ecological budgets and resource priors, and output the parameter set for the next scheduling cycle.

Images